摘要
针对以最小化总处理时间为目标的柔性车间调度问题,为优化工件设备加工路径,提出了求解强NP-Hard问题的改进人工免疫算法。在初始解产生方面应用了多种求解策略的组合,多个变异算子应用于工序分配和工序排序以产生新的抗体,能够有效保持种群的多样性。并通过引入重排序变异算子及克隆算子提高了算法的局部求解能力,使算法在局部求精与空间探索方面都取得了较好的成绩。通过在Bench Mark问题上的测试,并与相关文献仿真结果比较,表明了改进算法具有较好的稳定性和收敛性。
For the flexible job - hop scheduling problem targeting the total processing time minimization, an im- proved artificial immune algorithm is designed to solve the strong NP - Hard problem. The combination of multi - sol- ving strategies is applied in the initial solution; multiple mutation operator allocation processes and procedures are used in order to generate new antibodies, can effectively maintain the diversity of population. The improved artificial immune algorithm improves the local search ability by introducing the reordering mutation operator and clone operator. By using those operators, the proposed algorithm obtained better result in local refinement and space exploration. The algorithm is tested in the BenchMark problem, and comparison with several existing algorithms, to verify the effec- tiveness of the algorithm.
出处
《计算机仿真》
CSCD
北大核心
2014年第12期375-379,共5页
Computer Simulation
基金
国家自然科学基金(61203368)
关键词
人工免疫算法
变异算子
克隆
柔性车间调度问题
Artificial immune algorithm (AIA)
Mutation operator
Clone operator
Flexible job - hop scheduling problem (FJSP)
作者简介
马佳(1979-),女(汉族),辽宁盖州市人,博士,讲师,研究方向:智能控制、生产计划与调度、供应链管理的研究。
石刚(1978-),男(汉族),山东阳谷县人,博士,研究员,研究方向:人工智能、系统优化、模式识别等研究。